We are interested in knowing whether two-photon images can be further improved by computational means. Potential advantages are that images could be collected at lower signal-to-noise ratios (thereby minimizing exposure), and then adequate signal-to-noise ratio achieved by image restoration. Further, it is possible that improved resolution would result after image restoration, even for images collected at high signal-to-noise ratio. These possibilities are being tested by two-photon imaging of well characterized specimens to assess what improvements are possible.